Our brains have evolved to make decisions in changing environments, which means we have to continually learn in order to make adaptive choices. This type of learning requires two components: 1) maintaining a memory of how rewarding recent choices have been and 2) updating this memory following interactions with the environment (e.g., if a choice was better or worse than expected). While we know quite a bit about how we update our expectations, we know far less about where or how these memories are stored in the brain. As an M.D./Ph.D. student in Jeremiah Cohen’s lab, I have had the opportunity to study this problem and make inroads in understanding the neurobiology of decision-making. We took advantage of the mouse, a model organism increasingly used to study cognition. Using a combination of sophisticated behavioral paradigms, mathematical modeling, electrophysiology, pharmacology and optogenetics, we discovered that neurons in the mouse medial prefrontal cortex specifically maintained a memory of recent interactions with the environment. In particular, we found that individual neurons maintained this memory over long timescales. We also discovered a neural pathway that we believe provides a circuit mechanism for how these memories inform future choices. We believe this work answers a long-standing question of how the nervous system remembers interactions with the environment and allows for flexible decision-making. Given my long-term interests in understanding psychiatric disease, I think this work provides a rich substrate to explore disease and gain insight into how disorders of decision-making can be corrected.

In the Green lab, I developed a methodology to improve the current resolution of ribosome profiling technique, allowing us to systematically define the in vivo ribosome conformational states genome-wide. With this high-resolution ribosome profiling approach, we are able to decipher how cellular stresses regulate translation elongation.

I began my work with Tamara Lotan as a medical student at Johns Hopkins, and our research has focused on the biological determinants of racial disparities in prostate cancer. African-American men have a 1.5-fold higher incidence of prostate cancer than European-American men, and are nearly 2.5 times more likely to die from their disease. The biologic and genetic causes of these disparities are unknown. My work with Dr. Lotan has identified several molecular subtypes and genomic alterations of prostate cancer that are unique to African-American men. Importantly, some of these genomic alterations are directly associated with increased risks of metastasis and poor prognosis in this population. These genetic discoveries are critical for the development of prognostic and therapeutic approaches for precision medicine in African-American men.

This study is important because it brings to the forefront surgery and tumor biology in patients with colorectal cancer liver metastases. Specifically, it is the first study to tailor surgical technique (anatomical versus non-anatomical hepatectomies) according to the presence or absence of a biomarker (wild-type versus KRAS mutated tumors). Of note, to date, the use of biomarkers was limited to only informing prognosis. The broader clinical message of the study is that because tumor biology may differ between patients with the same malignancy, surgical treatment should be individualized.

In the lab of Gregg Semenza, my research interest is to investigate the role of hypoxia-inducible factors (HIFs) in the control of breast cancer stem cells (BCSCs) and to develop novel therapeutic strategies that will improve the response to chemotherapy by targeting BCSCs. We have shown that treatment with chemotherapy increases the BCSC population, which leads to drug resistance, tumor recurrence and metastasis. My studies have delineated HIF-regulated pathways that play important roles in the breast cancer cell response to chemotherapy. Targeting HIFs or downstream pathways will block induction of the BCSC phenotype and may improve clinical outcomes in breast cancer.

G-protein-coupled receptors (GPCRs) activate mTORC2-AKT signaling in metabolism, cell survival and cytoskeleton dynamics. Altered mTORC2-AKT signaling leads to many human diseases, including cancer, metastasis and metabolic syndromes; however, it is unknown how mTORC2 is activated downstream of GPCRs. Using proteomic approaches, I identified a GDP-bound RhoGTPase as an mTORC2-binding protein and identified its function in mTORC2-AKT signaling by reconstituting GPCR-mediated mTORC2-AKT activation with purified mTORC2, Rho and Ras. My discoveries rewrite the central dogma that G proteins are only active in a GTP-bound state and further our understanding of mTORC2-AKT-related human diseases.

In the Schneck and Mao labs, I engineered several biomaterials, including magnetic particles and extracellular matrix hydrogels, to overcome some of the challenges facing T-cell immunotherapies. In the process, we revealed key biology of T-cells such as nanoscale receptor organization and mechanical and environmental influences of T-cell activation, while extending our capacity to use these cancer-specific T-cells as a therapy.

Cells need to be in tune with their surroundings to decide when it is most favorable for them to grow. A protein called mTOR is key for sensing the availability of nutrients and serves as a switch between energy-producing and energy-consuming cellular processes. mTOR activity is enabled by its association with several proteins into the mTOR complex 1, and tight regulation of this activity is fundamental for the health of cells. Our research, conducted in the laboratory of Ted and Valina Dawson, has shown that another protein, called Thorase, can dismantle this mTOR complex 1. Mice that lack Thorase have an excess of mTOR activity, which leads to sudden death through a severe neurological condition reminiscent of what we observe in children harboring Thorase mutations. We have found that treating these mice with the mTOR inhibitor rapamycin greatly alleviates their disease. Since this drug and its derivatives are already used in the clinic, we hope that our findings will be readily translatable. Given that our lab has uncovered a broader role for Thorase in the protection of neurons, we believe our discovery will also have implications for other maladies, such as stroke.

Our brains encode information by using electrical signals (action potentials or spikes) within neurons, which then act at a circuit level to generate behavior. Neural coding is the study of how these spikes encode information, and can be broadly classified into “rate” and “temporal” coding. Rate coding refers to the simple idea that changes in the frequency of spiking carry information. In contrast, temporal coding is a concept that argues the temporal structure of spike sequence represents information. While temporal codes have been observed in a variety of systems, whether temporal codes alone are sufficient to drive behavior is controversial, and the molecular mechanisms underlying temporal codes are poorly defined. In Mark Wu’s lab at Johns Hopkins, I studied Drosophila circadian clock neurons and showed that their temporal pattern of firing directly controls sleep quality. Using genetic approaches, I also delineated the molecular mechanisms underlying these clock-driven spiking patterns. Surprisingly, my work also revealed that these temporal pattern codes can themselves trigger synaptic plasticity, the first time this has ever been shown.

Deepening our collective understanding of how blood vessels form has the potential to benefit the development of new therapies for two of the leading causes of death worldwide, cardiovascular disease and cancer. Interestingly, cardiovascular disease treatment aims to promote blood vessel formation, while cancer therapies aim to inhibit the formation of tumor vasculature, further highlighting the need to understand the details of this process in order to successfully control it in both contexts. Studying how human blood vessels form in the lab requires a highly biomimetic platform that can mimic aspects of the regenerative/pathological environment in which blood vessels form in vivo. To achieve this, I used oxygen-controllable hydrogels to study the impact of 3D hypoxic gradients on endothelial cell behavior. Hypoxia is a hallmark of pro-angiogenic environments and is known to regulate many vital biological processes. Using this system, I was able to recapitulate the key morphogenetic events of an understudied mechanism for blood vessel formation, identify regulators of the process, and control vascular networks by manipulating the environment in which the endothelial cells reside. I have done my research in the lab of Sharon Gerecht.

As a graduate student in the lab of Dr. Douglas N. Robinson, I have had the opportunity to study protein interactions that endow cells with the ability to sense and respond to the chemical and mechanical signals they experience.  Using the model organism Dictyostelium, we discovered that to ensure the cell’s ability to mount a quick, robust response to stimuli, a set of proteins critical for generating contractility form “contractility kits” in the cytoplasm.  Non-muscle myosin II, actin crosslinker cortexillin I, and scaffolding protein IQGAP2 build pre-formed units that are primed to respond to stimuli and accumulate to the actin cytoskeletal network, where they can relieve stress on the network and drive shape change.  Another concept we are revealing is that feedback exists between seemingly diverse processes in the cell, such as metabolism, transcription and translation, and cell mechanics.  By identifying the biochemical interactions that integrate and drive these processes, we are uncovering new biology that will help us further understand how cells sense and respond to their dynamic environment.

My work has been carried out in the Hunterian Neurosurgical Research Laboratory led by Henry Brem and Betty Tyler. While our laboratory is primarily focused on developing therapeutics for brain tumors, we saw an opportunity to expand our work to head and neck cancer in order to meet a significant clinical need. This project demonstrated that the FDA-approved antiviral drug ribavirin has a potent anti-tumor effect in vitro and in vivo in nasopharyngeal carcinoma (NPC), a malignancy with a propensity for metastasis and a prominent lack of available targeted therapies. We showed that ribavirin’s effects were mediated by modulating four protein targets known to play important roles in NPC cancer biology. Our work provides a foundation for clinical evaluation of ribavirin in NPC.

In the Goley lab, we study bacterial growth, development and division using Caulobacter crescentus as a model organism. During my thesis work, I discovered that the conserved transcriptional regulator CdnL regulates transcription of biosynthetic genes required for proliferation. Loss of CdnL alters the transcriptome in a manner that impacts metabolic homeostasis required for growth, morphogenesis and development. Consistent with CdnL’s role in proliferation, nutrient limitation leads to clearance of CdnL, indicating an undiscovered layer of control by which bacteria adapt to stress. These findings shed light on the intricate mechanisms bacteria use to regulate their growth and could inform medically relevant research aimed at developing tools to inhibit bacterial proliferation.

In Takanari Inoue’s lab, we use cell biology techniques to parse the signals that regulate physiological processes such as cell division, movement and fusion. All these events are triggered by mechanical forces that arise from actin protein polymerization on the cell membrane. To understand the signaling circuit that regulates this force, for the first time in our lab we set out to probe the signaling molecules outside a living cell, where we had more experimental control. We embarked on bottom-up engineering of an artificial cell-like system that could mimic actin-induced force generation on its membrane in response to a chemical input. We fabricated cell-sized vesicles and in them we reconstituted the chemical sensing modules, and actin polymerization regulators. By fine-tuning the concentrations of the proteins and lipids, we achieved actin-induced forces that broke symmetry in response to the placement of artificial cells in a chemical gradient. We then tracked the protein modules in time and space, and realized a bistable, feedback behavior. Our platform bridged the gap between cell studies and the in vitro ones that lack the cells’ dimensionality and membrane considerations. This protocell could serve as a biomimetic device that is deployable in a cellular milieu for drug or gene delivery.

The pathogenesis of Parkinson’s disease is due to accumulation and spreading of pathologic a-synuclein. However, what drives the abnormal assembly of pathologic a-synuclein and how dopaminergic neurons are dying by a-synuclein are not known. In our recent two papers, we found both neuronal and non-neuronal mechanism of a-synuclein-mediated neurodegeneration. Pathologic a-synuclein kills neurons via not only PARP-1 activation in neurons, but also releases neurotoxin from activated microglia and astrocyte.

I’ve always been told that practice makes perfect. In the developing auditory system, this also seems the case. Even before hearing begins, cells in the inner ear rehearse for hearing — sending electrical signals to the brain that are indistinguishable from pure tones. During my Ph.D. studies, I was able to visualize these signals in awake animals using fluorescent brain activity sensors. We found that brain activity in developing auditory centers is highly organized and propagates from the periphery all the way to the auditory cortex. This activity was not driven by sounds (the auditory canal is not open), but originated in the inner ear itself, as removal of the ears abolished this activity. While we don’t know the exact role this activity plays during development, we suspect that coordinated activation of brain cells across multiple brain areas helps refine and strengthen newly formed brain connections. Understanding how this activity influences development may help us gain new insights into auditory processing disorders in young children, the causes of which remain unknown and underexplored.

Metastasis is the primary driver of cancer-related deaths. Yet, its mechanisms remain poorly characterized. The current model of metastasis is based on loss of the cell-cell adhesion protein (E-cadherin) as a driver of cancer cell invasion and distant colonization; it is therefore classified as a tumor/ metastasis suppressor. My work directly demonstrates that, although E-cadherin suppresses invasion, its expression is required for successful metastasis in most breast cancers. I show that E-cadherin is a survival factor whose expression is essential for limiting oxidative-stress mediated cell death in cancer cells during metastasis.

We developed a super-achromatic flexible probe with an ultracompact form factor (approximately 520 micrometers in outer diameter), enabling ultrahigh-resolution (approximately 1.7 micrometers axial resolution (in tissue)) endoscopic optical coherence tomography imaging at 800 nanometers. This technology affords a great potential to perform “optical biopsies” without the need for tissue removal or processing. It has significant translational potential for in vivo clinical applications in assessing tissue pathological changes in internal luminal organs, particularly those of small and/or young subjects and complex internal organs (such as small airways). I am doing my research in the Biophotonics Imaging Technology Lab directed by Xingde Li at the Department of Biomedical Engineering.

I have the great opportunity to work in the laboratory of Peter Devreotes as a graduate student. The long-term goal of our lab is to have a complete description of the network controlling migratory behavior, which will provide new targets for drugs that could be applied for cell migration. Specifically, here we identified that cells maintain complementary spatial and temporal distributions of Ras activity and Phosphatidylinositol-3,4-Bisphosphate (PI(3,4)P2) during random migration and in response to chemoattractants. Whereas Ras is active in the front, PI(3,4)P2 is distributed in a back to front gradient. In addition, depletion of PI(3,4)P2 by disruption of the 5-phosphatase, Dd5P4, or by recruitment of 4-phosphatase INPP4B to the plasma membrane, leads to elevated Ras activity, cell spreading and altered migratory behavior. Taken together, my work investigated the molecular mechanisms that bring about the mutual inhibitory interaction between Ras activation and PI(3,4)P2. These exciting findings uncover an important role of PI(3,4)P2 in the regulation of Ras activity, which may extend well beyond cell migration.

Immune mediated loss of myelin sheaths (a process called demyelination) from the neurons in the brain and/or spinal cord causes neurological diseases such as multiple sclerosis (MS). In the human brain and spinal cord, myelin acts as an insulating material that coats and protects neurons and helps conduct information between the brain and various parts of our body. In order to improve function and reduce disability in patients with MS, it is desirable to develop therapies that both: 1) inhibit the activity of disease-causing immune cells, and 2) promote remyelination of axons by oligodendrocytes, the specialized cells that produce myelin sheath. Currently, all available drugs for MS focus on controlling a patient’s immune response. No drugs are available that can directly promote remyelination. In our laboratory, we have established an efficient method to grow human oligodendrocyte cells. We are using the gene editing technique called CRISPR/Cas9 to insert fluorescent markers and reporter sequences into specific genes so that the cells that become oligodendrocytes will be molecularly labeled. The labeled cells can be easily detected, identified and purified. Using these cells, we have established an assay platform to screen for drugs that promote myelin formation. We are beginning to test libraries of drugs to identify a lead drug that can promote the remyelination capacity of the human oligodendrocyte cells and prevent them from dying under stressful conditions. This research is conducted in Donald J. Zack’s laboratory in the Department of Ophthalmology.

Our body is made up of about 37 trillion cells. These cells communicate with each other to regulate homeostasis and sustain our life. For example, for neuronal communication, neurotransmitters are released from pre-synaptic neurons and received by post-synaptic neurons. During wound healing, epidermal growth factors are delivered to the neighboring cells for the regeneration of the damaged tissue. When cells are invaded by pathogens, messenger protein, interleukin, is immediately sent to immune cells. Receptor proteins for these signal transmitters are sitting on the surface of the cell (plasma membrane). To control the cellular communications, cells constantly alter their shapes by adding and removing membranes with the receptors to their surface. These membrane remodeling events essentially control all cellular functions and thus are fundamental to our life. However, how these processes are fueled is not well understood, particularly during endocytosis — the process by which plasma membrane is removed from the cell surface. During endocytosis, a piece of plasma membrane is invaginated and pinched off to generate a vesicle. The pinching-off reaction is mediated by the motor protein dynamin. Dynamin molecules form a ring-shaped structure at the base of the membrane invagination and constrict the membrane to close the gap between two opposing membranes. This mechanochemical reaction is fueled by intracellular energy currency guanosine triphosphate, GTP. Dynamin demands a large amount of GTP to generate the motive force, however the GTP concentration within cytoplasm is insufficient for the dynamin function. In contrast, the GTP concentration sufficient for dynamin function would be toxic to the cell. Then, how does the cell regulate GTP? To address this question, I decided to measure the GTP level during endocytosis using a fluorescence-based biosensor for GTP. To our surprise, the GTP level was transiently elevated at the site of endocytosis. This temporally elevated GTP-fueled dynamin activity. Furthermore, using a genetic approach, I identified a specific protein, which is recruited to the endocytic sites to generate GTP. These results demonstrate that cells regulate the GTP level locally and transiently during endocytosis and thereby have provided a new insight into the energy regulation for membrane remodeling. These findings have further implications in the field since GTP is consumed by molecular nanomachines used in many intracellular processes, such as synaptic vesicle recycling, proliferation of the cell, organelle homeostasis and cell migration. Thus, my research opened up a new avenue in the field to investigate local energy control for cellular functions. This research was done in the Shigeki Watanabe lab (Department of Cell Biology).

Despite the profound and durable clinical responses to checkpoint therapy that have led to FDA approval for PD1 and CTLA4 blockade in several tumor types, patients with prostate cancer have yet to benefit from these therapies. Understanding the immunosuppressive pathways underpinning the lack of anti-tumor responses in prostate cancer, as well as the mechanisms that regulate these pathways, may lead to novel treatment paradigms that unleash the potential of checkpoint therapy in the treatment of prostate cancer. As a Ph.D. student in Charles Drake’s laboratory, I found that an important immune-resistance mechanism is initiated as a byproduct of the main line of treatment for prostate cancer, androgen deprivation therapy (ADT). Androgen receptor loss of signaling following ADT induces prostate tumor cells to upregulate the expression of IL-8. IL-8 is a chemokine that recruits suppressive neutrophils in the context of cancer known as polymorphonuclear myeloid derived suppressor cells (PMN-MDSCs). Inhibiting IL-8 signaling led to a reduced recruitment of PMN-MDSCs to prostate tumors. Furthermore, hindering the recruitment of PMN-MDSCs in combination with checkpoint blockade significantly delayed tumor outgrowth. Our data provide a rationale for targeting PMN-MDSC recruitment in combination with immune checkpoint blockade for the treatment of prostate cancer before ADT is administrated, when the immune-suppressive microenvironment has not been yet established. Based on these findings, our group has launched an investigator-initiated clinical trial to evaluate targeting PMN-MDSCs in combination with checkpoint blockade before ADT in prostate cancer patients.